A tournament of linkage tests in complex inheritance

Area of Science:

• Genetics
• Statistical genetics
• Computational biology

Background:

• Accurate genetic linkage analysis is crucial for identifying genes underlying complex traits.
• Weakly parametric methods are commonly used but their performance can vary.
• Evaluating these methods on simulated data is essential for understanding their strengths and limitations.

Purpose of the Study:

• To compare the performance of commonly used weakly parametric linkage tests.
• To evaluate these tests under various sampling and variable conditions.
• To identify optimal methods for genetic linkage analysis in oligogenic inheritance.

Main Methods:

• Utilized 200 replicates of simulated oligogenic inheritance data from Genetic Analysis Workshop 10.
• Dichotomized quantitative traits at different thresholds and selected samples based on affected sibs, creating 8 data combinations.
• Compared variance component (SOLAR) and sib-pair (SIBPAL2, BETA) programs.

Main Results:

• SOLAR performed best with continuous traits, even in selected samples, using the population mean.
• SIBPAL2 was superior in most other scenarios, utilizing phenotype product, population mean, and empirical pair correlations.
• BETA showed slightly more power than maximum likelihood scores under the null hypothesis.

Conclusions:

• The choice of linkage analysis method depends on trait type and sampling strategy.
• Incorporating population mean and empirical correlations can enhance the performance of most methods.
• Further research is needed to determine if weakly parametric methods can fully account for ascertainment bias.